Fix HighMemoryUsage alert ratio

diagnostic/build-fb0df9c3.json

@@ -0,0 +1,90 @@
+{
+  "generated_at": "2026-07-09T12:11:02.292720+00:00",
+  "commit": "fb0df9c3",
+  "diagnostic_logd": [
+    "diagnostic\\build-fb0df9c3-part001.logd",
+    "diagnostic\\build-fb0df9c3-part002.logd",
+    "diagnostic\\build-fb0df9c3-part003.logd"
+  ],
+  "diagnostic_logd_error": null,
+  "chunked": true,
+  "chunk_size_bytes": 41943040,
+  "password": "a7430eb931d198efd7dd",
+  "decrypt_command": "encryptly unpack diagnostic\\build-fb0df9c3.logd <outdir> --password a7430eb931d198efd7dd",
+  "total_modules": 10,
+  "passed": 0,
+  "failed": 10,
+  "modules": [
+    {
+      "name": "backend",
+      "status": "FAIL",
+      "elapsed_seconds": 0,
+      "artifact": null,
+      "output": "Command not found: [WinError 2] \u7cfb\u7edf\u627e\u4e0d\u5230\u6307\u5b9a\u7684\u6587\u4ef6\u3002"
+    },
+    {
+      "name": "frontend",
+      "status": "FAIL",
+      "elapsed_seconds": 0.141,
+      "artifact": null,
+      "output": "npm install failed:\nnpm: can't open file 'C:\\\\Users\\\\27796\\\\Documents\\\\Codex\\\\2026-07-09\\\\github-bounty-pr-20-30-zhaoziyuan2024\\\\work\\\\zeroeye\\\\frontend\\\\install': [Errno 2] No such file or directory\n"
+    },
+    {
+      "name": "market",
+      "status": "FAIL",
+      "elapsed_seconds": 0,
+      "artifact": null,
+      "output": "Command not found: [WinError 2] \u7cfb\u7edf\u627e\u4e0d\u5230\u6307\u5b9a\u7684\u6587\u4ef6\u3002"
+    },
+    {
+      "name": "frailbox",
+      "status": "FAIL",
+      "elapsed_seconds": 0,
+      "artifact": null,
+      "output": "Command not found: [WinError 2] \u7cfb\u7edf\u627e\u4e0d\u5230\u6307\u5b9a\u7684\u6587\u4ef6\u3002"
+    },
+    {
+      "name": "engine",
+      "status": "FAIL",
+      "elapsed_seconds": 0,
+      "artifact": null,
+      "output": "Command not found: [WinError 2] \u7cfb\u7edf\u627e\u4e0d\u5230\u6307\u5b9a\u7684\u6587\u4ef6\u3002"
+    },
+    {
+      "name": "compliance",
+      "status": "FAIL",
+      "elapsed_seconds": 0,
+      "artifact": null,
+      "output": "Command not found: [WinError 2] \u7cfb\u7edf\u627e\u4e0d\u5230\u6307\u5b9a\u7684\u6587\u4ef6\u3002"
+    },
+    {
+      "name": "v2-market-stream",
+      "status": "FAIL",
+      "elapsed_seconds": 0,
+      "artifact": null,
+      "output": "Command not found: [WinError 2] \u7cfb\u7edf\u627e\u4e0d\u5230\u6307\u5b9a\u7684\u6587\u4ef6\u3002"
+    },
+    {
+      "name": "nfc-scanner",
+      "status": "FAIL",
+      "elapsed_seconds": 0,
+      "artifact": null,
+      "output": "Command not found: [WinError 2] \u7cfb\u7edf\u627e\u4e0d\u5230\u6307\u5b9a\u7684\u6587\u4ef6\u3002"
+    },
+    {
+      "name": "openapi-haskell",
+      "status": "FAIL",
+      "elapsed_seconds": 0,
+      "artifact": null,
+      "output": "Command not found: [WinError 2] \u7cfb\u7edf\u627e\u4e0d\u5230\u6307\u5b9a\u7684\u6587\u4ef6\u3002"
+    },
+    {
+      "name": "openapi-tools",
+      "status": "FAIL",
+      "elapsed_seconds": 0,
+      "artifact": null,
+      "output": "Command not found: [WinError 2] \u7cfb\u7edf\u627e\u4e0d\u5230\u6307\u5b9a\u7684\u6587\u4ef6\u3002"
+    }
+  ],
+  "pr_note": "Include the encrypted diagnostic logd artifact(s): diagnostic\\build-fb0df9c3-part001.logd, diagnostic\\build-fb0df9c3-part002.logd, diagnostic\\build-fb0df9c3-part003.logd. The encrypted .logd is the required diagnostic content for PR review; this JSON file is metadata. Maintainers may ask you to remove these diagnostic artifacts before merging."
+}

tests/__init__.py

@@ -0,0 +1 @@
+# Test package marker for unittest discovery.

tests/test_monitoring_setup.py

@@ -0,0 +1,67 @@
+import pathlib
+import sys
+import unittest
+
+
+ROOT = pathlib.Path(__file__).resolve().parents[1]
+sys.path.insert(0, str(ROOT))
+
+from tools import monitoring_setup
+
+
+class MonitoringSetupRuleTests(unittest.TestCase):
+    def test_recommended_alert_rules_do_not_self_divide(self):
+        self.assertEqual(
+            monitoring_setup.validate_alert_rules(
+                monitoring_setup.RECOMMENDED_ALERT_RULES
+            ),
+            [],
+        )
+
+    def test_high_memory_alert_uses_machine_memory_denominator(self):
+        high_memory_rule = next(
+            rule for rule in monitoring_setup.RECOMMENDED_ALERT_RULES
+            if rule["name"] == "HighMemoryUsage"
+        )
+
+        self.assertIn("process_resident_memory_bytes", high_memory_rule["expr"])
+        self.assertIn("machine_memory_bytes", high_memory_rule["expr"])
+        self.assertNotIn(
+            "process_resident_memory_bytes / process_resident_memory_bytes",
+            high_memory_rule["expr"],
+        )
+
+    def test_validation_rejects_self_dividing_alert_expression(self):
+        invalid_rules = [{
+            "name": "BadMemoryUsage",
+            "expr": (
+                "process_resident_memory_bytes{job='api'} / "
+                "process_resident_memory_bytes{job='api'} > 0.9"
+            ),
+        }]
+
+        self.assertEqual(
+            monitoring_setup.validate_alert_rules(invalid_rules),
+            [
+                "BadMemoryUsage: expression divides "
+                "process_resident_memory_bytes by itself"
+            ],
+        )
+
+    def test_dry_run_rejects_invalid_self_dividing_rule(self):
+        invalid_rules = [{
+            "name": "BadMemoryUsage",
+            "expr": "process_resident_memory_bytes / process_resident_memory_bytes > 0.9",
+        }]
+
+        self.assertFalse(
+            monitoring_setup.upload_prometheus_rules(
+                invalid_rules,
+                monitoring_setup.DEFAULT_PROMETHEUS_URL,
+                dry_run=True,
+            )
+        )
+
+
+if __name__ == "__main__":
+    unittest.main()

tools/monitoring_setup.py

@@ -25,6 +25,7 @@
 import argparse
 import json
 import os
+import re
 import sys
 import time
 import urllib.request
@@ -42,6 +43,14 @@
 
 DASHBOARD_DIR = os.path.join(os.path.dirname(__file__), "..", "monitoring", "dashboards")
 ALERT_RULES_DIR = os.path.join(os.path.dirname(__file__), "..", "monitoring", "alerts")
+METRIC_NAME_RE = r"[a-zA-Z_:][a-zA-Z0-9_:]*"
+SELF_DIVISION_RE = re.compile(
+    rf"\b({METRIC_NAME_RE})(?:\s*\{{[^{{}}]*\}})?\s*/\s*"
+    rf"(?:on\s*\([^)]*\)\s*)?"
+    rf"(?:ignoring\s*\([^)]*\)\s*)?"
+    rf"(?:group_(?:left|right)(?:\s*\([^)]*\))?\s*)?"
+    rf"({METRIC_NAME_RE})(?:\s*\{{[^{{}}]*\}})?\b"
+)
 
 RECOMMENDED_ALERT_RULES: List[Dict[str, Any]] = [
     {
@@ -78,7 +87,7 @@
     },
     {
         "name": "HighMemoryUsage",
-        "expr": "process_resident_memory_bytes / process_resident_memory_bytes > 0.9",
+        "expr": "process_resident_memory_bytes / on(instance) group_left machine_memory_bytes > 0.9",
         "duration": "10m",
         "severity": "warning",
         "summary": "High memory usage on {{$labels.instance}}",
@@ -146,12 +155,30 @@
     {"name": "job:http_requests_total:rate5m", "expr": "sum(rate(http_requests_total[5m])) by (job)"},
     {"name": "job:http_error_rate:rate5m", "expr": "sum(rate(http_errors_total[5m])) by (job) / sum(rate(http_requests_total[5m])) by (job)"},
     {"name": "job:http_latency_p99:rate5m", "expr": "histogram_quantile(0.99, sum(rate(http_request_duration_seconds_bucket[5m])) by (le, job))"},
-    {"name": "instance:memory_usage:ratio", "expr": "process_resident_memory_bytes / machine_memory_bytes"},
+    {"name": "instance:memory_usage:ratio", "expr": "process_resident_memory_bytes / on(instance) group_left machine_memory_bytes"},
     {"name": "instance:cpu_usage:ratio", "expr": "rate(process_cpu_seconds_total[5m])"},
     {"name": "service:uptime:days", "expr": "time() - process_start_time_seconds{job=~'.+'}"},
 ]
 
 
+def find_self_dividing_expressions(rules: List[Dict[str, Any]]) -> List[str]:
+    """Return rule names whose expressions divide the same metric by itself."""
+    errors = []
+    for rule in rules:
+        expr = str(rule.get("expr", ""))
+        for lhs_metric, rhs_metric in SELF_DIVISION_RE.findall(expr):
+            if lhs_metric == rhs_metric:
+                name = rule.get("name", "<unnamed>")
+                errors.append(
+                    f"{name}: expression divides {lhs_metric} by itself"
+                )
+    return errors
+
+
+def validate_alert_rules(rules: List[Dict[str, Any]]) -> List[str]:
+    return find_self_dividing_expressions(rules)
+
+
 def http_request(method: str, url: str, data: Any = None,
                  headers: Optional[Dict[str, str]] = None) -> Any:
     if headers is None:
@@ -200,6 +227,13 @@ def check_alertmanager(url: str) -> bool:
 def upload_prometheus_rules(rules: List[Dict[str, Any]],
                             prometheus_url: str,
                             dry_run: bool = False) -> bool:
+    validation_errors = validate_alert_rules(rules)
+    if validation_errors:
+        print("Invalid Prometheus alert rules:", file=sys.stderr)
+        for error in validation_errors:
+            print(f"  - {error}", file=sys.stderr)
+        return False
+
     rules_file = "/etc/prometheus/rules/tent_rules.yml"
     print(f"{'Would upload' if dry_run else 'Uploading'} {len(rules)} rules to {prometheus_url}")
 
@@ -414,13 +448,14 @@ def main():
                 args.alertmanager_url, args.slack_webhook,
                 args.pagerduty_key, args.dry_run)
 
-        upload_prometheus_rules(RECOMMENDED_ALERT_RULES, args.prometheus_url, args.dry_run)
+        if not upload_prometheus_rules(RECOMMENDED_ALERT_RULES, args.prometheus_url, args.dry_run):
+            return 1
         print("Monitoring initialization complete")
         return 0
 
     if args.alerts:
-        upload_prometheus_rules(RECOMMENDED_ALERT_RULES, args.prometheus_url, args.dry_run)
-        return 0
+        return 0 if upload_prometheus_rules(
+            RECOMMENDED_ALERT_RULES, args.prometheus_url, args.dry_run) else 1
 
     if args.backup:
         return 0 if backup_monitoring_config(